Lancet Draft Output

---

##########################################

# extract counts and proportions for plotting

# composite reference_df
agg_reference_df <- dplyr::left_join(
  total_counts %>% 
    dplyr::filter(Year >= 2011) %>%
    dplyr::group_by(Year) %>%
    dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                     health_references = sum(health_count, na.rm = T),
                     intersection_references = sum(intersection_count, na.rm = T)
    ) %>%
    tidyr::pivot_longer(cols = -c("Year"),
                        names_to = "Type", values_to = "Count"),
  total_counts %>% 
    dplyr::filter(Year >= 2011) %>%
    dplyr::group_by(Year) %>%
    dplyr::summarize(climate_references = mean(climate_count, na.rm = T),
                     health_references = mean(health_count, na.rm = T),
                     intersection_references = mean(intersection_count, na.rm = T)
    ) %>%
    tidyr::pivot_longer(cols = -c("Year"),
                        names_to = "Type", values_to = "Avg")) %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L))

# health sector reference_df

reference_df <- dplyr::left_join(
  total_counts %>% 
    dplyr::filter(Year >= 2011) %>%
    dplyr::group_by(Sector, Year) %>%
    dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                     health_references = sum(health_count, na.rm = T),
                     intersection_references = sum(intersection_count, na.rm = T)
    ) %>%
    tidyr::pivot_longer(cols = -c("Sector", "Year"),
                        names_to = "Type", values_to = "Count"),
  total_counts %>% 
    dplyr::filter(Year >= 2011) %>%
    dplyr::group_by(Sector, Year) %>%
    dplyr::summarize(climate_references = mean(climate_count, na.rm = T),
                     health_references = mean(health_count, na.rm = T),
                     intersection_references = mean(intersection_count, na.rm = T)
    ) %>%
    tidyr::pivot_longer(cols = -c("Sector", "Year"),
                        names_to = "Type", values_to = "Avg")) %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L))

# composite plot_df
agg_plot_df <- total_texts %>% dplyr::filter(Year >= 2011)  %>%
  tidyr::pivot_longer(cols = -c("Sector", "Year", "total_texts"),
                      names_to = "text", values_to = "value") %>%
  dplyr::group_by(Year, text) %>%
  dplyr::summarise(total_texts = sum(total_texts, na.rm = T),
                   value = sum(value, na.rm = T)) %>%
  dplyr::mutate(value = tidyr::replace_na(value, 0),
                Prop = value/total_texts,
                Key = factor(dplyr::case_when(text == "climate_texts" ~ "Climate",
                                              text == "intersection_texts" ~ "Intersection",
                                              text == "health_texts" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L))

# health sector plot_df
plot_df <- total_texts %>% dplyr::filter(Year >= 2011)  %>%
  tidyr::pivot_longer(cols = -c("Sector", "Year", "total_texts"),
                      names_to = "text", values_to = "value") %>%
  dplyr::mutate(value = tidyr::replace_na(value, 0),
                Prop = value/total_texts,
                Key = factor(dplyr::case_when(text == "climate_texts" ~ "Climate",
                                              text == "intersection_texts" ~ "Intersection",
                                              text == "health_texts" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L))

---

Data summary

data_summary <- tibble::tibble(
  count = length(total_counts$Participant),
  unique = unique(total_counts$Participant) %>% length(),
  total_health = total_counts %>% dplyr::filter(!is.na(health_count)) %>% dplyr::filter(health_count != 0) %>% nrow(),
  total_climate = total_counts %>% dplyr::filter(!is.na(climate_count)) %>% dplyr::filter(climate_count != 0) %>% nrow(),
  total_intersection = total_counts %>% dplyr::filter(!is.na(intersection_count)) %>% dplyr::filter(intersection_count != 0) %>% nrow()
) %>%
  dplyr::mutate(prop_health = round(total_health/count,2),
                prop_climate = round(total_climate/count,2),
                prop_intersection = round(total_intersection/count,2)) 

colnames(data_summary) <- c("Companies (N)", "Companies (Unique)", 
                             "Health, (N)" , "Climate, (N)", "Intersection, (N)", 
                             "Health, %" , "Climate, %", "Intersection, %")
data_summary %>%
  knitr::kable()

Companies (N)	Companies (Unique)	Health, (N)	Climate, (N)	Intersection, (N)	Health, %	Climate, %	Intersection, %
17984	5329	14528	12297	4627	0.81	0.68	0.26

readr::write_csv(data_summary, "../output/0_0_data_summary.csv")

---

Yearly breakdown

a <- total_counts %>% dplyr::filter(Year >= 2011) %>% dplyr::group_by(Year) %>% 
  dplyr::summarize("Companies (N)" = n_distinct(Participant)) 

b <- agg_plot_df %>%
  dplyr::select(Year, Key, value) %>%
  dplyr::mutate(Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "value") %>%
  dplyr::rename("Climate (N)" = "Climate", "Health (N)" = "Health", "Intersection (N)" = "Intersection")
  
c <- agg_plot_df %>%
  dplyr::select(Year, Key, Prop) %>%
  dplyr::mutate(Prop = round(Prop, 2),
                Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Prop") %>%
  dplyr::rename("Climate (Prop)" = "Climate", "Health (Prop)" = "Health", "Intersection (Prop)" = "Intersection")
  
yearly_breakdown <- dplyr::left_join(a, b) %>% left_join(., c) 

yearly_breakdown %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Companies (N)	Climate (N)	Health (N)	Intersection (N)	Climate (Prop)	Health (Prop)	Intersection (Prop)
2011	1021	773	868	282	0.73	0.82	0.27
2012	1374	998	1150	327	0.70	0.81	0.23
2013	1553	1069	1287	343	0.67	0.80	0.21
2014	1686	1131	1369	357	0.66	0.79	0.21
2015	1771	1194	1463	408	0.66	0.81	0.22
2016	1928	1266	1546	440	0.64	0.78	0.22
2017	1972	1352	1594	484	0.67	0.79	0.24
2018	2000	1392	1638	553	0.68	0.80	0.27
2019	2197	1568	1864	637	0.70	0.83	0.28
2020	2029	1547	1742	791	0.75	0.84	0.38

readr::write_csv(yearly_breakdown, "../output/0_1_yearly_breakdown.csv")

---

Table of English language COP reports by year

reports_by_year <- total_counts %>% dplyr::filter(Year >= 2011) %>% dplyr::group_by(Year) %>% 
  dplyr::summarize(count = n_distinct(Participant)) %>%
  dplyr::rename("Companies (N)" = "count")

reports_by_year %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Companies (N)
2011	1021
2012	1374
2013	1553
2014	1686
2015	1771
2016	1928
2017	1972
2018	2000
2019	2197
2020	2029

readr::write_csv(reports_by_year, "../output/0_2_reports_by_year.csv")

All Sectors (Composite)

These are the figures generated with the composite measures from all sectors

a. Main text - Proportion of companies, %

Plot

## a. Main text - Proportion of companies, %
p <- agg_plot_df %>% 
  ggplot(aes(x = Year, y = Prop, color = Key)) +
  geom_path(aes(linetype = Key)) +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2013-05-31"), y = 0.85, label = "Health", color = "#619cff") +
  ggplot2::annotate("text", x = as.Date("2013-05-31"), y = 0.48, label = "Climate Change", color = "darkgreen") +
  ggplot2::annotate("text", x = as.Date("2012-12-31"), y = 0.1, label = "Intersection", color = "red") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0,1)) +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Proportion of companies, %") 

p

ggsave("../output/1a_prop_of_companies.pdf", p,  width = 10, height = 7)

Table

t <- agg_plot_df %>%
  dplyr::select(Year, Key, Prop) %>%
  dplyr::mutate(Prop = round(Prop, 2),
                Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Prop") 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Climate	Health	Intersection
2011	0.73	0.82	0.27
2012	0.70	0.81	0.23
2013	0.67	0.80	0.21
2014	0.66	0.79	0.21
2015	0.66	0.81	0.22
2016	0.64	0.78	0.22
2017	0.67	0.79	0.24
2018	0.68	0.80	0.27
2019	0.70	0.83	0.28
2020	0.75	0.84	0.38

readr::write_csv(t, "../output/1a_prop_of_companies.csv")

---

b. Appendix - Total number of references

Plot

## b. Appendix - Total number of references
p <- agg_reference_df %>%
  ggplot(aes(x = Year, y = Count, color = Key)) +
  geom_line(aes(linetype = Key))  +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2011-05-31"), y = 31500, label = "Health", color = "#619cff") +
  ggplot2::annotate("text", x = as.Date("2013-05-31"), y = 25000, label = "Climate Change", color = "darkgreen") +
  ggplot2::annotate("text", x = as.Date("2016-12-31"), y = 4000, label = "Intersection", color = "red") +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Total number of references") 

p

ggsave("../output/1b_number_of_references.pdf", p,  width = 10, height = 7)

Table

t <- agg_plot_df %>%
  dplyr::select(Year, Key, value) %>%
  dplyr::mutate(Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "value") 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Climate	Health	Intersection
2011	773	868	282
2012	998	1150	327
2013	1069	1287	343
2014	1131	1369	357
2015	1194	1463	408
2016	1266	1546	440
2017	1352	1594	484
2018	1392	1638	553
2019	1568	1864	637
2020	1547	1742	791

readr::write_csv(t, "../output/1b_number_of_references.csv")

---

c. Appendix - Total number of references (Intersection)

Plot

## c. Appendix - Total number of references (Intersection)
p <- agg_reference_df %>% 
  dplyr::filter(Key == "Intersection") %>%
  ggplot(aes(x = Year, y = Count, color = Key)) +
  geom_line() +
  ggplot2::annotate("text", x = as.Date("2013-12-31"), y = 1500, label = "Intersection", color = "red") +
  theme_minimal() +
  scale_y_continuous(limits = c(0, 4000)) +
  theme(legend.position = "none") +
  labs(y = "Total number of references") 

p

ggsave("../output/1c_number_of_references_intersection.pdf", p,  width = 10, height = 7)

Table

t <- agg_reference_df %>%
  dplyr::select(Year, Key, Count) %>%
  dplyr::mutate(Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Count") %>%
  dplyr::select(Year, Intersection) 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Intersection
2011	1023
2012	989
2013	1320
2014	1321
2015	1305
2016	1571
2017	1739
2018	1828
2019	2421
2020	3365

readr::write_csv(t, "../output/1c_number_of_references_intersection.csv")

---

d. Appendix - Proportion of companies, % (Intersection)

Plot

p <- agg_plot_df %>% 
  dplyr::filter(Key == "Intersection") %>%
  ggplot(aes(x = Year, y = Prop, color = Key)) +
  geom_path(aes(linetype = Key)) +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2015-12-31"), y = 0.3, label = "Intersection", color = "red") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 0.4)) +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Proportion of companies, %") 

p

ggsave("../output/1d_proportion_intersection.pdf", p,  width = 10, height = 7)

Table

agg_plot_df %>%
  dplyr::select(Year, Key, Prop) %>%
  dplyr::mutate(Year = lubridate::year(Year),
                Prop = round(Prop, 2)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Prop") %>%
  dplyr::select(Year, Intersection) 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Intersection
2011	1023
2012	989
2013	1320
2014	1321
2015	1305
2016	1571
2017	1739
2018	1828
2019	2421
2020	3365

readr::write_csv(t, "../output/1d_proportion_intersection.csv")

---

e. Appendix - Average number of references

Plot

p <- agg_reference_df %>%
  ggplot(aes(x = Year, y = Avg, color = Key)) +
  geom_line(aes(linetype = Key)) +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2011-05-31"), y = 29, label = "Health", color = "#619cff") +
  ggplot2::annotate("text", x = as.Date("2014-01-01"), y = 17, label = "Climate Change", color = "darkgreen") +
  ggplot2::annotate("text", x = as.Date("2016-12-31"), y = 3, label = "Intersection", color = "red") +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Average number of references") 

p

ggsave("../output/1e_avg_references.pdf", p,  width = 10, height = 7)

Table

t <- agg_reference_df %>%
  dplyr::select(Year, Key, Avg) %>%
  dplyr::mutate(Year = lubridate::year(Year),
                Avg = round(Avg,2)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Avg")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Climate	Health	Intersection
2011	20.26	26.96	1.10
2012	17.06	24.99	0.80
2013	15.05	24.05	0.96
2014	15.46	25.95	0.90
2015	13.87	24.89	0.84
2016	14.41	24.19	0.95
2017	14.53	24.55	1.02
2018	15.21	25.98	1.05
2019	17.48	26.14	1.24
2020	22.94	32.98	1.84

readr::write_csv(t, "../output/1e_avg_references.csv")

---

f. Appendix - Total number of references by WHO region

Plot

## f. Appendix - Total number of references by WHO region
p <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::group_by(Year, who_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "who_region"),
                      names_to = "Type", values_to = "Count") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection") %>%
  ggplot(aes(x = Year, y = Count, color = who_region)) +
  geom_path() +
  theme_minimal() +
  labs(y = "Total number of references",
       color = "WHO Region") 

p

ggsave("../output/1f_who_number_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::group_by(Year, who_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::arrange(who_region, Year) 

colnames(t) <- c("Year", "WHO Region", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	WHO Region	Climate	Health	Intersection
2011	Africa	444	788	56
2012	Africa	242	960	20
2013	Africa	429	1460	38
2014	Africa	552	1536	73
2015	Africa	663	1880	50
2016	Africa	686	1988	41
2017	Africa	486	1383	51
2018	Africa	397	1210	31
2019	Africa	551	1527	52
2020	Africa	459	907	42
2011	Eastern Mediterranean	124	475	14
2012	Eastern Mediterranean	236	945	18
2013	Eastern Mediterranean	201	1133	29
2014	Eastern Mediterranean	196	721	13
2015	Eastern Mediterranean	290	1083	10
2016	Eastern Mediterranean	398	1073	29
2017	Eastern Mediterranean	349	1048	28
2018	Eastern Mediterranean	793	1662	46
2019	Eastern Mediterranean	748	1675	47
2020	Eastern Mediterranean	478	1450	51
2011	Europe	9455	12354	525
2012	Europe	10235	15910	547
2013	Europe	10684	15754	657
2014	Europe	12147	18610	680
2015	Europe	11079	16718	626
2016	Europe	12369	18716	803
2017	Europe	12299	19129	836
2018	Europe	12988	20714	891
2019	Europe	16866	24566	1217
2020	Europe	21566	28534	1597
2011	Latin America and the Caribbean	563	1214	41
2012	Latin America and the Caribbean	1339	2128	52
2013	Latin America and the Caribbean	876	1703	60
2014	Latin America and the Caribbean	804	1719	65
2015	Latin America and the Caribbean	703	1661	15
2016	Latin America and the Caribbean	958	1656	38
2017	Latin America and the Caribbean	1306	1961	64
2018	Latin America and the Caribbean	1179	2514	102
2019	Latin America and the Caribbean	1092	2837	76
2020	Latin America and the Caribbean	1756	3554	152
2011	North America	1755	3479	125
2012	North America	1853	3416	123
2013	North America	1589	4206	152
2014	North America	1840	5760	134
2015	North America	2188	7050	253
2016	North America	2341	6677	265
2017	North America	2553	6623	281
2018	North America	2538	6043	247
2019	North America	3562	5848	311
2020	North America	3978	6842	398
2011	South-East Asia	824	1527	57
2012	South-East Asia	868	1769	56
2013	South-East Asia	1201	2353	76
2014	South-East Asia	937	2272	55
2015	South-East Asia	1158	3050	105
2016	South-East Asia	1415	2648	116
2017	South-East Asia	944	2820	111
2018	South-East Asia	1073	2642	110
2019	South-East Asia	1230	2911	128
2020	South-East Asia	1123	2448	166
2011	Western Pacific	5755	5341	205
2012	Western Pacific	6308	5757	173
2013	Western Pacific	5641	6341	308
2014	Western Pacific	6097	7262	301
2015	Western Pacific	5489	7263	246
2016	Western Pacific	5579	7100	279
2017	Western Pacific	6748	8739	368
2018	Western Pacific	7476	10397	401
2019	Western Pacific	10069	11653	590
2020	Western Pacific	12474	16419	959

readr::write_csv(t, "../output/1f_who_number_references.csv")

---

g. Appendix - Proportion of references by WHO region

Plot

## j. Appendix - Total number of regerences by by SIDS, Tier 1 and Tier 2 categories

p <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, who_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "who_region"),
                      names_to = "Type", values_to = "Prop") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_prop" ~ "Climate",
                                              Type == "intersection_prop" ~ "Intersection",
                                              Type == "health_prop" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection" & !is.na(who_region)) %>%
  ggplot(aes(x = Year, y = Prop, color = who_region)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Proportion of companies, %",
       color = "") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 0.5))

p

ggsave("../output/1g_who_prop_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, who_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  dplyr::arrange(desc(who_region), Year) %>%
  dplyr::filter(!is.na(who_region)) 

colnames(t) <- c("Year", "WHO Region", "Total documents", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	WHO Region	Total documents	Climate	Health	Intersection
2011	Western Pacific	165	0.84	0.87	0.38
2012	Western Pacific	215	0.78	0.84	0.30
2013	Western Pacific	222	0.77	0.86	0.34
2014	Western Pacific	225	0.76	0.85	0.35
2015	Western Pacific	225	0.74	0.86	0.36
2016	Western Pacific	248	0.73	0.82	0.30
2017	Western Pacific	257	0.79	0.86	0.37
2018	Western Pacific	265	0.79	0.85	0.41
2019	Western Pacific	300	0.77	0.86	0.41
2020	Western Pacific	279	0.84	0.91	0.57
2011	South-East Asia	91	0.62	0.76	0.24
2012	South-East Asia	107	0.59	0.79	0.20
2013	South-East Asia	129	0.63	0.82	0.22
2014	South-East Asia	134	0.55	0.83	0.16
2015	South-East Asia	156	0.58	0.78	0.21
2016	South-East Asia	151	0.56	0.79	0.23
2017	South-East Asia	145	0.57	0.86	0.23
2018	South-East Asia	146	0.55	0.84	0.20
2019	South-East Asia	156	0.59	0.86	0.28
2020	South-East Asia	132	0.58	0.86	0.26
2011	North America	101	0.76	0.83	0.30
2012	North America	133	0.70	0.83	0.26
2013	North America	139	0.65	0.78	0.26
2014	North America	141	0.64	0.82	0.19
2015	North America	153	0.67	0.82	0.26
2016	North America	156	0.67	0.84	0.27
2017	North America	160	0.79	0.86	0.37
2018	North America	167	0.73	0.83	0.33
2019	North America	188	0.76	0.84	0.33
2020	North America	181	0.87	0.90	0.46
2011	Latin America and the Caribbean	61	0.56	0.66	0.23
2012	Latin America and the Caribbean	104	0.55	0.62	0.23
2013	Latin America and the Caribbean	95	0.55	0.62	0.22
2014	Latin America and the Caribbean	106	0.48	0.60	0.16
2015	Latin America and the Caribbean	82	0.54	0.66	0.13
2016	Latin America and the Caribbean	137	0.42	0.46	0.14
2017	Latin America and the Caribbean	133	0.50	0.50	0.19
2018	Latin America and the Caribbean	143	0.49	0.50	0.21
2019	Latin America and the Caribbean	127	0.48	0.50	0.23
2020	Latin America and the Caribbean	133	0.56	0.60	0.32
2011	Europe	557	0.75	0.82	0.25
2012	Europe	743	0.74	0.82	0.22
2013	Europe	863	0.71	0.81	0.19
2014	Europe	950	0.70	0.80	0.21
2015	Europe	997	0.69	0.80	0.22
2016	Europe	1068	0.68	0.79	0.22
2017	Europe	1100	0.69	0.79	0.22
2018	Europe	1095	0.72	0.82	0.26
2019	Europe	1236	0.73	0.84	0.27
2020	Europe	1181	0.77	0.85	0.37
2011	Eastern Mediterranean	42	0.48	0.81	0.19
2012	Eastern Mediterranean	63	0.54	0.76	0.22
2013	Eastern Mediterranean	80	0.40	0.78	0.11
2014	Eastern Mediterranean	77	0.44	0.70	0.06
2015	Eastern Mediterranean	114	0.51	0.80	0.08
2016	Eastern Mediterranean	118	0.50	0.80	0.12
2017	Eastern Mediterranean	121	0.50	0.83	0.09
2018	Eastern Mediterranean	136	0.51	0.82	0.18
2019	Eastern Mediterranean	131	0.56	0.84	0.19
2020	Eastern Mediterranean	99	0.58	0.82	0.24
2011	Africa	42	0.71	0.93	0.21
2012	Africa	61	0.52	0.87	0.13
2013	Africa	75	0.45	0.85	0.13
2014	Africa	93	0.47	0.81	0.13
2015	Africa	90	0.48	0.91	0.16
2016	Africa	100	0.56	0.89	0.17
2017	Africa	94	0.50	0.83	0.18
2018	Africa	89	0.53	0.87	0.18
2019	Africa	113	0.55	0.88	0.17
2020	Africa	61	0.61	0.85	0.25

readr::write_csv(t, "../output/1g_who_prop_references.csv")

---

h. Appendix - Total number of regerences by SIDS, Tier 1 and Tier 2

Plot

## i. Appendix - Total number of regerences by SIDS, Tier 1 and Tier 2
p <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::group_by(Year, tier_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "tier_region"),
                      names_to = "Type", values_to = "Count") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection" & !is.na(tier_region)) %>%
  ggplot(aes(x = Year, y = Count, color = tier_region)) +
  geom_path() +
  theme_minimal() +
  labs(y = "Total number of references",
       color = "") 

p

ggsave("../output/1h_sids_number_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::group_by(Year, tier_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::filter(!is.na(tier_region)) %>%
  dplyr::arrange(tier_region, Year)

colnames(t) <- c("Year", "Region", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Region	Climate	Health	Intersection
2011	SIDS	274	455	21
2012	SIDS	428	579	18
2013	SIDS	350	842	40
2014	SIDS	376	618	34
2015	SIDS	350	607	29
2016	SIDS	203	357	32
2017	SIDS	258	561	18
2018	SIDS	417	748	47
2019	SIDS	676	1085	63
2020	SIDS	793	1288	105
2011	Tier 1	1657	3461	126
2012	Tier 1	1799	3477	126
2013	Tier 1	1590	4118	138
2014	Tier 1	2021	6124	136
2015	Tier 1	2287	7343	263
2016	Tier 1	2415	6739	276
2017	Tier 1	2485	6446	285
2018	Tier 1	2469	6014	222
2019	Tier 1	2937	5411	290
2020	Tier 1	3662	6332	370
2011	Tier 2	4973	7638	369
2012	Tier 2	4602	9329	285
2013	Tier 2	5696	9828	493
2014	Tier 2	5930	9485	322
2015	Tier 2	5516	9505	314
2016	Tier 2	6400	10721	432
2017	Tier 2	6292	10744	466
2018	Tier 2	6358	10888	571
2019	Tier 2	8451	13223	573
2020	Tier 2	9395	13899	729

readr::write_csv(t, "../output/1h_sids_number_references.csv")

---

i. Appendix - Proportion of references by SIDS, Tier 1 and Tier 2

Plot

## j. Appendix - Total number of regerences by by SIDS, Tier 1 and Tier 2 categories

p <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, tier_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "tier_region"),
                      names_to = "Type", values_to = "Prop") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_prop" ~ "Climate",
                                              Type == "intersection_prop" ~ "Intersection",
                                              Type == "health_prop" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection" & !is.na(tier_region)) %>%
  ggplot(aes(x = Year, y = Prop, color = tier_region)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Proportion of companies, %",
       color = "") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 0.5))

p

ggsave("../output/1i_sids_prop_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, tier_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  dplyr::arrange(desc(tier_region), Year) %>%
  dplyr::filter(!is.na(tier_region)) 

colnames(t) <- c("Year", "Category", "Total documents", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Category	Total documents	Climate	Health	Intersection
2011	Tier 2	296	0.76	0.79	0.28
2012	Tier 2	374	0.75	0.79	0.20
2013	Tier 2	439	0.69	0.81	0.23
2014	Tier 2	462	0.68	0.81	0.19
2015	Tier 2	492	0.66	0.77	0.21
2016	Tier 2	518	0.68	0.79	0.24
2017	Tier 2	509	0.70	0.80	0.22
2018	Tier 2	513	0.70	0.78	0.27
2019	Tier 2	571	0.74	0.82	0.27
2020	Tier 2	522	0.75	0.83	0.36
2011	Tier 1	107	0.70	0.78	0.28
2012	Tier 1	144	0.66	0.80	0.26
2013	Tier 1	145	0.65	0.80	0.28
2014	Tier 1	150	0.65	0.79	0.21
2015	Tier 1	156	0.67	0.82	0.26
2016	Tier 1	168	0.67	0.83	0.26
2017	Tier 1	165	0.78	0.85	0.33
2018	Tier 1	179	0.70	0.81	0.30
2019	Tier 1	195	0.71	0.82	0.30
2020	Tier 1	184	0.83	0.90	0.42
2011	SIDS	16	0.81	0.94	0.31
2012	SIDS	24	0.71	0.88	0.25
2013	SIDS	30	0.67	0.77	0.30
2014	SIDS	27	0.52	0.78	0.30
2015	SIDS	34	0.56	0.79	0.21
2016	SIDS	30	0.57	0.77	0.20
2017	SIDS	29	0.66	0.90	0.38
2018	SIDS	34	0.74	0.94	0.38
2019	SIDS	50	0.74	0.96	0.36
2020	SIDS	45	0.73	0.89	0.49

readr::write_csv(t, "../output/1i_sids_prop_references.csv")

---

j. Appendix - Total number of regerences by HDI categories

Plot

## j. Appendix - Total number of regerences by HDI categories
p <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::group_by(Year, hdi_level) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "hdi_level"),
                      names_to = "Type", values_to = "Count") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L),
                hdi_level = factor(hdi_level, levels = c("Very high human development",
                                                         "High human development",
                                                         "Medium human development",
                                                         "Low human development"))) %>%
  dplyr::filter(Key == "Intersection" & !is.na(hdi_level)) %>%
  ggplot(aes(x = Year, y = Count, color = hdi_level)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Total number of references",
       color = "") 

p

ggsave("../output/1j_hdi_number_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>% 
  dplyr::group_by(Year, hdi_level) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::arrange(desc(hdi_level), Year) %>%
  dplyr::filter(!is.na(hdi_level))

colnames(t) <- c("Year", "HDI Level", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	HDI Level	Climate	Health	Intersection
2011	Very high human development	16600	20564	820
2012	Very high human development	18267	24889	833
2013	Very high human development	17807	26430	1106
2014	Very high human development	19867	31121	1094
2015	Very high human development	18624	30789	1097
2016	Very high human development	20204	32259	1333
2017	Very high human development	21492	34297	1465
2018	Very high human development	22757	36927	1531
2019	Very high human development	30285	41851	2100
2020	Very high human development	37984	51885	2939
2011	Medium human development	707	1318	60
2012	Medium human development	900	2034	52
2013	Medium human development	736	1879	65
2014	Medium human development	634	1900	74
2015	Medium human development	815	2376	95
2016	Medium human development	961	2526	96
2017	Medium human development	653	2261	77
2018	Medium human development	734	2382	91
2019	Medium human development	916	2290	80
2020	Medium human development	682	1641	115
2011	Low human development	29	131	5
2012	Low human development	42	133	3
2013	Low human development	13	148	0
2014	Low human development	10	113	0
2015	Low human development	30	175	7
2016	Low human development	53	155	3
2017	Low human development	56	143	10
2018	Low human development	140	398	16
2019	Low human development	149	472	13
2020	Low human development	28	122	13
2011	High human development	1568	3129	136
2012	High human development	1853	3793	101
2013	High human development	2015	4261	126
2014	High human development	2017	4633	138
2015	High human development	2071	5250	98
2016	High human development	2524	4890	137
2017	High human development	2460	4976	186
2018	High human development	2781	5443	188
2019	High human development	2767	6402	228
2020	High human development	3122	6478	296

readr::write_csv(t, "../output/1j_hdi_number_references.csv")

---

k. Appendix - Proportion of references by HDI categories

Plot

## j. Appendix - Total number of regerences by HDI categories

p <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, hdi_level) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "hdi_level"),
                      names_to = "Type", values_to = "Prop") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_prop" ~ "Climate",
                                              Type == "intersection_prop" ~ "Intersection",
                                              Type == "health_prop" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L),
                hdi_level = factor(hdi_level, levels = c("Very high human development",
                                                         "High human development",
                                                         "Medium human development",
                                                         "Low human development"))) %>%
  dplyr::filter(Key == "Intersection" & !is.na(hdi_level)) %>%
  ggplot(aes(x = Year, y = Prop, color = hdi_level)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Proportion of companies, %",
       color = "") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 0.5))

p

ggsave("../output/1k_hdi_prop_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011) %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, hdi_level) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  dplyr::arrange(desc(hdi_level), Year) %>%
  dplyr::filter(!is.na(hdi_level)) 

colnames(t) <- c("Year", "HDI Level", "Total documents", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	HDI Level	Total documents	Climate	Health	Intersection
2011	Very high human development	795	0.77	0.84	0.28
2012	Very high human development	1065	0.75	0.82	0.24
2013	Very high human development	1207	0.71	0.81	0.22
2014	Very high human development	1304	0.70	0.80	0.23
2015	Very high human development	1369	0.70	0.81	0.25
2016	Very high human development	1472	0.68	0.80	0.24
2017	Very high human development	1521	0.72	0.80	0.26
2018	Very high human development	1540	0.73	0.82	0.30
2019	Very high human development	1741	0.74	0.84	0.30
2020	Very high human development	1654	0.79	0.86	0.41
2011	Medium human development	101	0.52	0.78	0.21
2012	Medium human development	136	0.54	0.85	0.20
2013	Medium human development	142	0.51	0.81	0.13
2014	Medium human development	148	0.51	0.80	0.11
2015	Medium human development	172	0.48	0.79	0.15
2016	Medium human development	172	0.54	0.81	0.16
2017	Medium human development	173	0.53	0.81	0.15
2018	Medium human development	166	0.52	0.84	0.15
2019	Medium human development	175	0.55	0.83	0.20
2020	Medium human development	123	0.54	0.84	0.20
2011	Low human development	9	0.56	0.78	0.22
2012	Low human development	13	0.38	0.77	0.08
2013	Low human development	24	0.17	0.71	0.00
2014	Low human development	25	0.20	0.64	0.00
2015	Low human development	24	0.46	0.79	0.12
2016	Low human development	28	0.36	0.68	0.11
2017	Low human development	29	0.31	0.76	0.10
2018	Low human development	29	0.45	0.79	0.24
2019	Low human development	39	0.46	0.79	0.15
2020	Low human development	19	0.47	0.63	0.21
2011	High human development	153	0.65	0.73	0.24
2012	High human development	211	0.58	0.70	0.21
2013	High human development	229	0.57	0.79	0.23
2014	High human development	248	0.54	0.76	0.17
2015	High human development	250	0.55	0.78	0.15
2016	High human development	304	0.52	0.70	0.18
2017	High human development	285	0.55	0.74	0.21
2018	High human development	301	0.55	0.70	0.21
2019	High human development	295	0.54	0.74	0.24
2020	High human development	267	0.59	0.76	0.30

readr::write_csv(t, "../output/1k_hdi_prop_references.csv")

---

l. Appendix - Number of references 2020 per sector

## f. Appendix - Number of references 2020 per sector
t <- total_counts %>% 
  dplyr::filter(Year == 2020) %>%
  dplyr::group_by(Sector, Year) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::select(Sector, health_references, climate_references, intersection_references) 

colnames(t) <- c("Sector", "Health", "Climate", "Intersection")

readr::write_csv(t, "../output/1l_sector_references_2020.csv")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Health	Climate	Intersection
Aerospace & Defense	754	403	46
Alternative Energy	551	866	92
Automobiles & Parts	2032	2160	105
Banks	1079	1484	84
Beverages	927	562	37
Chemicals	3438	2078	191
Construction & Materials	3357	2699	192
Diversified	1789	1024	93
Electricity	1184	2041	77
Electronic & Electrical Equ…	1605	822	64
Equity Investment Instruments	80	99	12
Financial Services	3217	4243	210
Fixed Line Telecommunications	327	195	24
Food & Drug Retailers	244	195	16
Food Producers	2907	1282	202
Forestry & Paper	367	580	22
Gas, Water & Multiutilities	921	821	91
General Industrials	4532	2836	274
General Retailers	1470	1216	40
Health Care Equipment & Ser…	1751	356	45
Household Goods & Home Cons…	1072	1022	48
Industrial Engineering	1158	767	82
Industrial Goods & Services	12	4	0
Industrial Metals & Mining	1358	789	74
Industrial Transportation	1275	819	51
Leisure Goods	136	63	2
Life Insurance	651	330	41
Media	680	454	19
Mining	1150	594	55
Mobile Telecommunications	925	629	61
Nonequity Investment Instru…	112	37	3
Nonlife Insurance	413	260	23
Not Applicable	29	16	0
Oil & Gas Producers	1603	1490	114
Oil Equipment, Services & D…	594	435	24
Personal Goods	806	504	22
Pharmaceuticals & Biotechno…	6336	1125	280
Real Estate Investment & Se…	1638	1289	103
Real Estate Investment Trusts	232	241	24
Software & Computer Services	1404	1098	106
Support Services	3662	2174	200
Technology Hardware & Equip…	1632	1131	99
Travel & Leisure	744	601	17

---

m. Appendix - Average of references 2020 per sector

## g. Appendix - Average of references 2020 per sector
t <- total_counts %>% 
  dplyr::filter(Year == 2020) %>%
  dplyr::group_by(Sector, Year) %>%
  dplyr::summarize(climate_references = mean(climate_count, na.rm = T),
                   health_references = mean(health_count, na.rm = T),
                   intersection_references = mean(intersection_count, na.rm = T)
  ) %>%
  dplyr::mutate_if(is.numeric,round,1) %>%
  dplyr::select(Sector, health_references, climate_references, intersection_references) 


colnames(t) <-c("Sector", "Health, %", "Climate, %", "Intersection, %")

readr::write_csv(t, "../output/1m_avg_sector_references_2020.csv")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Health, %	Climate, %	Intersection, %
Aerospace & Defense	47.1	25.2	2.9
Alternative Energy	24.0	37.7	4.0
Automobiles & Parts	48.4	51.4	2.5
Banks	21.6	29.7	1.7
Beverages	29.0	17.6	1.2
Chemicals	58.3	35.2	3.2
Construction & Materials	41.4	33.3	2.4
Diversified	19.7	11.3	1.0
Electricity	34.8	60.0	2.3
Electronic & Electrical Equ…	26.8	13.7	1.1
Equity Investment Instruments	8.9	11.0	1.3
Financial Services	24.4	32.1	1.6
Fixed Line Telecommunications	27.2	16.2	2.0
Food & Drug Retailers	61.0	48.8	4.0
Food Producers	37.3	16.4	2.6
Forestry & Paper	30.6	48.3	1.8
Gas, Water & Multiutilities	41.9	37.3	4.1
General Industrials	32.4	20.3	2.0
General Retailers	28.3	23.4	0.8
Health Care Equipment & Ser…	42.7	8.7	1.1
Household Goods & Home Cons…	35.7	34.1	1.6
Industrial Engineering	30.5	20.2	2.2
Industrial Goods & Services	6.0	2.0	0.0
Industrial Metals & Mining	52.2	30.3	2.8
Industrial Transportation	28.3	18.2	1.1
Leisure Goods	15.1	7.0	0.2
Life Insurance	65.1	33.0	4.1
Media	18.9	12.6	0.5
Mining	63.9	33.0	3.1
Mobile Telecommunications	37.0	25.2	2.4
Nonequity Investment Instru…	56.0	18.5	1.5
Nonlife Insurance	31.8	20.0	1.8
Not Applicable	29.0	16.0	0.0
Oil & Gas Producers	53.4	49.7	3.8
Oil Equipment, Services & D…	24.8	18.1	1.0
Personal Goods	17.5	11.0	0.5
Pharmaceuticals & Biotechno…	124.2	22.1	5.5
Real Estate Investment & Se…	35.6	28.0	2.2
Real Estate Investment Trusts	29.0	30.1	3.0
Software & Computer Services	15.4	12.1	1.2
Support Services	18.3	10.9	1.0
Technology Hardware & Equip…	32.6	22.6	2.0
Travel & Leisure	22.5	18.2	0.5

---

n. Appendix - Proportion of companies 2020 per sector

Plot

## h. Appendix - Proportion of companies 2020 per sector
p <- plot_df %>%
  dplyr::filter(Year >= "2020-01-01") %>% 
  ggplot(., aes(x = Sector, y = Prop, fill = Key)) +
  geom_bar(stat = "identity", position = "dodge") +
  theme_minimal() +
  scale_y_continuous(labels = scales::percent) +
  theme(axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1)) +
  theme(legend.position = "bottom") +
  labs(fill = "",
       y = "Proportion of companies, %",
       x = "\nSector")

p

ggsave("../output/1n_prop_companies_sector.pdf", p,  width = 10, height = 7)

Table

t <- plot_df %>%
  dplyr::filter(Year >= "2020-01-01") %>%
  dplyr::select(Sector, Year, Key, Prop) %>%
  dplyr::mutate(Year = lubridate::year(Year), 
                Prop = round(Prop, 2)) %>%
  dplyr::arrange(Sector, Year) %>%
  tidyr::pivot_wider(names_from = "Key", values_from = "Prop") 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Year	Climate	Health	Intersection
Aerospace & Defense	2020	0.72	0.89	0.39
Alternative Energy	2020	0.91	0.96	0.70
Automobiles & Parts	2020	0.83	0.89	0.55
Banks	2020	0.87	0.85	0.42
Beverages	2020	0.61	0.86	0.33
Chemicals	2020	0.85	0.89	0.68
Construction & Materials	2020	0.78	0.83	0.44
Diversified	2020	0.78	0.84	0.31
Electricity	2020	0.79	0.79	0.51
Electronic & Electrical Equ…	2020	0.66	0.85	0.29
Equity Investment Instruments	2020	0.60	0.90	0.30
Financial Services	2020	0.80	0.77	0.37
Fixed Line Telecommunications	2020	0.69	0.85	0.38
Food & Drug Retailers	2020	1.00	1.00	0.75
Food Producers	2020	0.76	0.89	0.47
Forestry & Paper	2020	0.71	0.86	0.36
Gas, Water & Multiutilities	2020	0.76	0.88	0.60
General Industrials	2020	0.77	0.86	0.39
General Retailers	2020	0.71	0.81	0.31
Health Care Equipment & Ser…	2020	0.70	0.89	0.37
Household Goods & Home Cons…	2020	0.74	0.94	0.45
Industrial Engineering	2020	0.67	0.73	0.33
Industrial Goods & Services	2020	0.50	1.00	0.00
Industrial Metals & Mining	2020	0.89	0.89	0.46
Industrial Transportation	2020	0.73	0.81	0.33
Leisure Goods	2020	0.64	0.73	0.18
Life Insurance	2020	0.91	0.64	0.36
Media	2020	0.73	0.75	0.14
Mining	2020	0.89	0.89	0.58
Mobile Telecommunications	2020	0.81	0.96	0.50
Nonequity Investment Instru…	2020	0.50	1.00	0.50
Nonlife Insurance	2020	0.77	1.00	0.38
Not Applicable	2020	1.00	1.00	0.00
Oil & Gas Producers	2020	0.84	0.91	0.72
Oil Equipment, Services & D…	2020	0.67	0.89	0.37
Personal Goods	2020	0.70	0.81	0.22
Pharmaceuticals & Biotechno…	2020	0.82	0.93	0.60
Real Estate Investment & Se…	2020	0.78	0.92	0.50
Real Estate Investment Trusts	2020	0.70	0.80	0.60
Software & Computer Services	2020	0.63	0.83	0.25
Support Services	2020	0.66	0.81	0.27
Technology Hardware & Equip…	2020	0.83	0.85	0.36
Travel & Leisure	2020	0.76	0.79	0.16

readr::write_csv(t, "../output/1n_prop_companies_sector.csv")

---

o. Table with COVID and Gender dictionary counts in intersection documents

intersection_covid %>% knitr::kable(col.names = c("Year", "Total documents", "Keyword hits", "Flagged documents (N)", "Flagged documents (Prop)")) %>%
  kableExtra::add_header_above(c("", "", "COVID dictionary" = 3)) %>% 
  kableExtra::kable_styling()

		COVID dictionary
Year	Total documents	Keyword hits	Flagged documents (N)	Flagged documents (Prop)
2020	791	128	113	0.1428571

intersection_gender %>% knitr::kable(col.names = c("Year", "Total documents", "Keyword hits", "Flagged documents (N)", "Flagged documents (Prop)")) %>%
  kableExtra::add_header_above(c("", "", "Gender dictionary" = 3)) %>% 
  kableExtra::kable_styling()

		Gender dictionary
Year	Total documents	Keyword hits	Flagged documents (N)	Flagged documents (Prop)
2011	282	26	24	0.09
2012	327	20	17	0.05
2013	343	29	21	0.06
2014	357	26	17	0.05
2015	408	37	27	0.07
2016	440	51	40	0.09
2017	484	57	48	0.10
2018	553	71	59	0.11
2019	637	151	123	0.19
2020	791	113	99	0.13

readr::write_csv(intersection_covid, "../output/1o_intersection_covid.csv")
readr::write_csv(intersection_gender, "../output/1o_intersection_gender.csv")

p <- ggplot(intersection_gender, aes(x = year, y = prop_doct, group = 1)) +
  geom_line(color = "#cc0055") +
  theme_minimal() +
  labs(x= "Year",
       y = "Proportion of documents\n Gender mention in intersection, %\n") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 0.25))

p

ggsave("../output/1o_prop_intersecion_gender.pdf", p,  width = 10, height = 7)

---

Health Care Sector

These are the figures generated with the Health Care Equipment and Services Sector

p. Main text - Proportion of companies, % (Health Care Sector)

Plot

## a. Main text - Proportion of companies, %
p <- plot_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser...") %>%
  ggplot(aes(x = Year, y = Prop, color = Key)) +
  geom_line(aes(linetype = Key)) +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2013-05-31"), y = 0.85, label = "Health", color = "#619cff") +
  ggplot2::annotate("text", x = as.Date("2013-05-31"), y = 0.48, label = "Climate Change", color = "darkgreen") +
  ggplot2::annotate("text", x = as.Date("2012-12-31"), y = 0.1, label = "Intersection", color = "red") +
  scale_y_continuous(labels = scales::percent_format()) +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Proportion of companies, %") 

p

ggsave("../output/2p_prop_of_companies.pdf", p,  width = 10, height = 7)

Table

t <- plot_df %>%
  dplyr::filter(Sector == "Health Care Equipment & Ser...") %>%
  dplyr::select(Year, Key, Prop) %>%
  dplyr::mutate(Prop = round(Prop, 2),
                Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Prop")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Climate	Health	Intersection
2011	0.57	1.00	0.00
2012	0.50	0.89	0.22
2013	0.34	0.78	0.16
2014	0.33	0.80	0.13
2015	0.47	0.87	0.17
2016	0.50	0.78	0.19
2017	0.56	0.85	0.09
2018	0.73	0.89	0.35
2019	0.69	0.95	0.31
2020	0.70	0.89	0.37

readr::write_csv(t, "../output/2p_prop_of_companies.csv")

---

q. Appendix - Total number of references (Health Care Sector)

Plot

## b. Appendix - Total number of references
p <- reference_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser...") %>%
  ggplot(aes(x = Year, y = Count, color = Key)) +
  geom_line(aes(linetype = Key))  +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2011-05-31"), y = 500, label = "Health", color = "#619cff") +
  ggplot2::annotate("text", x = as.Date("2013-05-31"), y = 250, label = "Climate Change", color = "darkgreen") +
  ggplot2::annotate("text", x = as.Date("2016-12-31"), y = 75, label = "Intersection", color = "red") +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Total number of references") 

p

ggsave("../output/2q_number_of_references.pdf", p,  width = 10, height = 7)

Table

t <- reference_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser...") %>%
  dplyr::select(Year, Key, Count) %>%
  dplyr::mutate(Year = lubridate::year(Year)) %>%
  tidyr::pivot_wider(id_cols = Year, names_from = "Key", values_from = "Count") 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Climate	Health	Intersection
2011	90	377	0
2012	82	478	7
2013	139	908	40
2014	152	1248	66
2015	195	1277	27
2016	162	891	27
2017	279	1112	5
2018	314	1870	48
2019	352	1979	45
2020	356	1751	45

readr::write_csv(t, "../output/2q_number_of_references.csv")

---

r. Appendix - Total number of references (Intersection) — (Health Care Sector)

Plot

## c. Appendix - Total number of references (Intersection)
p <- reference_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser..." & Key == "Intersection") %>%
  ggplot(aes(x = Year, y = Count, color = Key)) +
  geom_line() +
  ggplot2::annotate("text", x = as.Date("2013-12-31"), y = 18, label = "Intersection", color = "red") +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Total number of references") 

p

ggsave("../output/2r_number_of_references_intersection.pdf", p,  width = 10, height = 7)

Table

t <- reference_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser..." & Key == "Intersection") %>%
  dplyr::mutate(Year = lubridate::year(Year)) %>%
  dplyr::select(Sector, Year, Count) 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Year	Count
Health Care Equipment & Ser…	2011	0
Health Care Equipment & Ser…	2012	7
Health Care Equipment & Ser…	2013	40
Health Care Equipment & Ser…	2014	66
Health Care Equipment & Ser…	2015	27
Health Care Equipment & Ser…	2016	27
Health Care Equipment & Ser…	2017	5
Health Care Equipment & Ser…	2018	48
Health Care Equipment & Ser…	2019	45
Health Care Equipment & Ser…	2020	45

readr::write_csv(t, "../output/2r_number_of_references_intersection.csv")

---

s. Appendix - Average number of references (Health Care Sector)

Plot

## d. Appendix - Average number of references
p <- reference_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser...") %>%
  ggplot(aes(x = Year, y = Avg, color = Key)) +
  geom_line(aes(linetype = Key)) +
  scale_linetype_manual(values = c("solid", "longdash", "twodash")) +
  ggplot2::annotate("text", x = as.Date("2011-05-31"), y = 32, label = "Health", color = "#619cff") +
  ggplot2::annotate("text", x = as.Date("2012-05-31"), y = 8, label = "Climate Change", color = "darkgreen") +
  ggplot2::annotate("text", x = as.Date("2016-12-31"), y = 3, label = "Intersection", color = "red") +
  theme_minimal() +
  theme(legend.position = "none") +
  labs(y = "Average number of references") 

p

ggsave("../output/2s_avg_references.pdf", p,  width = 10, height = 7)

Table

t <- reference_df %>% 
  dplyr::filter(Sector == "Health Care Equipment & Ser..." & Key == "Intersection") %>%
  dplyr::mutate(Year = lubridate::year(Year),
                Avg = round(Avg, 2)) %>%
  dplyr::select(Sector, Year, Avg) 

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Year	Avg
Health Care Equipment & Ser…	2011	0.00
Health Care Equipment & Ser…	2012	0.44
Health Care Equipment & Ser…	2013	1.60
Health Care Equipment & Ser…	2014	2.75
Health Care Equipment & Ser…	2015	1.04
Health Care Equipment & Ser…	2016	1.00
Health Care Equipment & Ser…	2017	0.17
Health Care Equipment & Ser…	2018	1.37
Health Care Equipment & Ser…	2019	1.12
Health Care Equipment & Ser…	2020	1.10

readr::write_csv(t, "../output/2s_avg_references.csv")

---

t. Appendix - Total number of references by WHO region (Health Care Sector)

Plot

## e. Appendix - Total number of references by WHO region
p <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::group_by(Sector, Year, who_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  tidyr::pivot_longer(cols = -c("Sector", "Year", "who_region"),
                      names_to = "Type", values_to = "Count") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection") %>%
  ggplot(aes(x = Year, y = Count, color = who_region)) +
  geom_path() +
  theme_minimal() +
  labs(y = "Total number of references",
       color = "WHO Region") 

p

ggsave("../output/2t_who_number_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::group_by(Sector, Year, who_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::arrange(who_region, Year) 

colnames(t) <- c("Sector", "Year", "WHO Region", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Year	WHO Region	Climate	Health	Intersection
Health Care Equipment & Ser…	2011	Africa	1	48	0
Health Care Equipment & Ser…	2012	Africa	3	115	1
Health Care Equipment & Ser…	2013	Africa	0	86	0
Health Care Equipment & Ser…	2014	Africa	24	201	44
Health Care Equipment & Ser…	2015	Africa	32	175	12
Health Care Equipment & Ser…	2016	Africa	2	53	2
Health Care Equipment & Ser…	2017	Africa	0	9	0
Health Care Equipment & Ser…	2018	Africa	8	107	1
Health Care Equipment & Ser…	2019	Africa	33	144	3
Health Care Equipment & Ser…	2020	Africa	19	68	2
Health Care Equipment & Ser…	2011	Eastern Mediterranean	0	12	0
Health Care Equipment & Ser…	2012	Eastern Mediterranean	0	10	0
Health Care Equipment & Ser…	2013	Eastern Mediterranean	9	195	0
Health Care Equipment & Ser…	2014	Eastern Mediterranean	0	2	0
Health Care Equipment & Ser…	2015	Eastern Mediterranean	0	3	0
Health Care Equipment & Ser…	2019	Eastern Mediterranean	0	57	0
Health Care Equipment & Ser…	2020	Eastern Mediterranean	0	6	0
Health Care Equipment & Ser…	2011	Europe	19	67	0
Health Care Equipment & Ser…	2012	Europe	28	114	0
Health Care Equipment & Ser…	2013	Europe	49	245	8
Health Care Equipment & Ser…	2014	Europe	71	608	10
Health Care Equipment & Ser…	2015	Europe	87	436	4
Health Care Equipment & Ser…	2016	Europe	78	455	4
Health Care Equipment & Ser…	2017	Europe	127	519	1
Health Care Equipment & Ser…	2018	Europe	157	1026	25
Health Care Equipment & Ser…	2019	Europe	189	760	21
Health Care Equipment & Ser…	2020	Europe	220	1016	34
Health Care Equipment & Ser…	2012	Latin America and the Caribbean	15	117	3
Health Care Equipment & Ser…	2013	Latin America and the Caribbean	5	122	1
Health Care Equipment & Ser…	2014	Latin America and the Caribbean	0	0	0
Health Care Equipment & Ser…	2015	Latin America and the Caribbean	3	163	0
Health Care Equipment & Ser…	2016	Latin America and the Caribbean	0	0	0
Health Care Equipment & Ser…	2017	Latin America and the Caribbean	0	0	0
Health Care Equipment & Ser…	2018	Latin America and the Caribbean	3	10	0
Health Care Equipment & Ser…	2019	Latin America and the Caribbean	7	267	0
Health Care Equipment & Ser…	2020	Latin America and the Caribbean	3	187	0
Health Care Equipment & Ser…	2011	North America	2	8	0
Health Care Equipment & Ser…	2012	North America	2	8	1
Health Care Equipment & Ser…	2013	North America	0	10	0
Health Care Equipment & Ser…	2014	North America	0	0	0
Health Care Equipment & Ser…	2017	North America	29	42	2
Health Care Equipment & Ser…	2018	North America	25	70	10
Health Care Equipment & Ser…	2019	North America	35	447	13
Health Care Equipment & Ser…	2020	North America	19	97	1
Health Care Equipment & Ser…	2011	South-East Asia	2	44	0
Health Care Equipment & Ser…	2012	South-East Asia	0	4	0
Health Care Equipment & Ser…	2013	South-East Asia	0	0	0
Health Care Equipment & Ser…	2014	South-East Asia	0	112	0
Health Care Equipment & Ser…	2015	South-East Asia	0	170	0
Health Care Equipment & Ser…	2016	South-East Asia	10	43	0
Health Care Equipment & Ser…	2017	South-East Asia	10	47	0
Health Care Equipment & Ser…	2018	South-East Asia	10	46	0
Health Care Equipment & Ser…	2019	South-East Asia	10	61	0
Health Care Equipment & Ser…	2020	South-East Asia	6	8	2
Health Care Equipment & Ser…	2011	Western Pacific	66	198	0
Health Care Equipment & Ser…	2012	Western Pacific	34	110	2
Health Care Equipment & Ser…	2013	Western Pacific	76	250	31
Health Care Equipment & Ser…	2014	Western Pacific	57	325	12
Health Care Equipment & Ser…	2015	Western Pacific	73	330	11
Health Care Equipment & Ser…	2016	Western Pacific	72	340	21
Health Care Equipment & Ser…	2017	Western Pacific	113	495	2
Health Care Equipment & Ser…	2018	Western Pacific	111	611	12
Health Care Equipment & Ser…	2019	Western Pacific	78	243	8
Health Care Equipment & Ser…	2020	Western Pacific	89	369	6

readr::write_csv(t, "../output/2t_who_number_references.csv")

---

u. Appendix - Proportion references by WHO region (Health Care Sector)

Plot

p <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, who_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "who_region"),
                      names_to = "Type", values_to = "Prop") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_prop" ~ "Climate",
                                              Type == "intersection_prop" ~ "Intersection",
                                              Type == "health_prop" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection" & !is.na(who_region)) %>%
  ggplot(aes(x = Year, y = Prop, color = who_region)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Proportion of companies, %",
       color = "") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 1))

p

ggsave("../output/2u_who_prop_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, who_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  dplyr::arrange(desc(who_region), Year) %>%
  dplyr::filter(!is.na(who_region)) 

colnames(t) <- c("Year", "WHO Region", "Total documents", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	WHO Region	Total documents	Climate	Health	Intersection
2011	Western Pacific	3	1.00	1.00	0.00
2012	Western Pacific	1	1.00	1.00	1.00
2013	Western Pacific	5	0.80	0.80	0.40
2014	Western Pacific	3	0.67	0.67	0.33
2015	Western Pacific	2	1.00	1.00	0.50
2016	Western Pacific	5	1.00	0.80	0.60
2017	Western Pacific	5	0.60	0.60	0.20
2018	Western Pacific	5	1.00	1.00	0.40
2019	Western Pacific	5	0.80	0.80	0.40
2020	Western Pacific	4	1.00	1.00	0.25
2011	South-East Asia	2	0.50	1.00	0.00
2012	South-East Asia	1	0.00	1.00	0.00
2013	South-East Asia	1	0.00	0.00	0.00
2014	South-East Asia	2	0.00	1.00	0.00
2015	South-East Asia	2	0.00	1.00	0.00
2016	South-East Asia	2	0.50	1.00	0.00
2017	South-East Asia	2	0.50	1.00	0.00
2018	South-East Asia	2	0.50	1.00	0.00
2019	South-East Asia	2	0.50	1.00	0.00
2020	South-East Asia	2	1.00	1.00	0.50
2011	North America	1	1.00	1.00	0.00
2012	North America	2	0.50	1.00	0.50
2013	North America	3	0.00	0.67	0.00
2014	North America	1	0.00	0.00	0.00
2017	North America	1	1.00	1.00	1.00
2018	North America	2	1.00	1.00	1.00
2019	North America	3	1.00	1.00	0.67
2020	North America	4	1.00	1.00	0.25
2012	Latin America and the Caribbean	2	0.50	0.50	0.50
2013	Latin America and the Caribbean	3	0.33	0.33	0.33
2014	Latin America and the Caribbean	1	0.00	0.00	0.00
2015	Latin America and the Caribbean	5	0.20	0.40	0.00
2016	Latin America and the Caribbean	2	0.00	0.00	0.00
2017	Latin America and the Caribbean	2	0.00	0.00	0.00
2018	Latin America and the Caribbean	4	0.50	0.25	0.00
2019	Latin America and the Caribbean	3	0.67	0.67	0.00
2020	Latin America and the Caribbean	3	0.67	0.67	0.00
2011	Europe	5	0.40	1.00	0.00
2012	Europe	10	0.50	0.90	0.00
2013	Europe	16	0.31	0.88	0.12
2014	Europe	18	0.39	0.83	0.11
2015	Europe	15	0.60	0.93	0.20
2016	Europe	19	0.47	0.79	0.11
2017	Europe	21	0.67	0.95	0.05
2018	Europe	20	0.80	0.95	0.40
2019	Europe	22	0.82	1.00	0.36
2020	Europe	29	0.66	0.86	0.45
2011	Eastern Mediterranean	2	0.00	1.00	0.00
2012	Eastern Mediterranean	1	0.00	1.00	0.00
2013	Eastern Mediterranean	3	0.33	1.00	0.00
2014	Eastern Mediterranean	1	0.00	1.00	0.00
2015	Eastern Mediterranean	1	0.00	1.00	0.00
2019	Eastern Mediterranean	1	0.00	1.00	0.00
2020	Eastern Mediterranean	1	0.00	1.00	0.00
2011	Africa	1	1.00	1.00	0.00
2012	Africa	1	1.00	1.00	1.00
2013	Africa	1	0.00	1.00	0.00
2014	Africa	4	0.25	1.00	0.25
2015	Africa	5	0.40	1.00	0.20
2016	Africa	4	0.25	1.00	0.25
2017	Africa	3	0.00	1.00	0.00
2018	Africa	4	0.25	1.00	0.25
2019	Africa	6	0.17	1.00	0.17
2020	Africa	3	0.33	1.00	0.33

readr::write_csv(t, "../output/2u_who_prop_references.csv")

---

v. Appendix - Total number of regerences by SIDS, Tier 1 and Tier 2 (Health Care Sector)

Plot

## i. Appendix - Total number of regerences by SIDS, Tier 1 and Tier 2
p <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::group_by(Sector, Year, tier_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  tidyr::pivot_longer(cols = -c("Sector", "Year", "tier_region"),
                      names_to = "Type", values_to = "Count") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection" & !is.na(tier_region)) %>%
  ggplot(aes(x = Year, y = Count, color = tier_region)) +
  geom_path() +
  theme_minimal() +
  labs(y = "Total number of references",
       color = "") 

p

ggsave("../output/2v_sids_number_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::group_by(Sector, Year, tier_region) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::filter(!is.na(tier_region)) %>%
  dplyr::arrange(tier_region, Year) 

colnames(t) <- c("Sector", "Year", "Region", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Year	Region	Climate	Health	Intersection
Health Care Equipment & Ser…	2018	SIDS	1	8	0
Health Care Equipment & Ser…	2019	SIDS	4	10	2
Health Care Equipment & Ser…	2011	Tier 1	2	8	0
Health Care Equipment & Ser…	2012	Tier 1	2	8	1
Health Care Equipment & Ser…	2013	Tier 1	2	16	0
Health Care Equipment & Ser…	2014	Tier 1	0	0	0
Health Care Equipment & Ser…	2016	Tier 1	2	0	0
Health Care Equipment & Ser…	2017	Tier 1	29	42	2
Health Care Equipment & Ser…	2018	Tier 1	25	70	10
Health Care Equipment & Ser…	2019	Tier 1	35	447	13
Health Care Equipment & Ser…	2020	Tier 1	19	97	1
Health Care Equipment & Ser…	2011	Tier 2	19	44	0
Health Care Equipment & Ser…	2012	Tier 2	35	180	3
Health Care Equipment & Ser…	2013	Tier 2	39	234	2
Health Care Equipment & Ser…	2014	Tier 2	49	428	10
Health Care Equipment & Ser…	2015	Tier 2	55	353	1
Health Care Equipment & Ser…	2016	Tier 2	44	197	2
Health Care Equipment & Ser…	2017	Tier 2	39	197	0
Health Care Equipment & Ser…	2018	Tier 2	69	420	21
Health Care Equipment & Ser…	2019	Tier 2	120	543	14
Health Care Equipment & Ser…	2020	Tier 2	42	521	14

readr::write_csv(t, "../output/2v_sids_number_references.csv")

---

w. Appendix - Proportion references by by SIDS, Tier 1 and Tier 2 (Health Care Sector)

Plot

p <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, tier_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "tier_region"),
                      names_to = "Type", values_to = "Prop") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_prop" ~ "Climate",
                                              Type == "intersection_prop" ~ "Intersection",
                                              Type == "health_prop" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L)) %>%
  dplyr::filter(Key == "Intersection" & !is.na(tier_region)) %>%
  ggplot(aes(x = Year, y = Prop, color = tier_region)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Proportion of companies, %",
       color = "") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 1))

p

ggsave("../output/2w_sids_prop_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, tier_region) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  dplyr::arrange(desc(tier_region), Year) %>%
  dplyr::filter(!is.na(tier_region)) 

colnames(t) <- c("Year", "Category", "Total documents", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	Category	Total documents	Climate	Health	Intersection
2011	Tier 2	3	0.67	1.00	0.00
2012	Tier 2	6	0.67	0.83	0.17
2013	Tier 2	11	0.36	0.64	0.18
2014	Tier 2	10	0.50	0.80	0.20
2015	Tier 2	12	0.50	0.75	0.08
2016	Tier 2	8	0.50	0.50	0.12
2017	Tier 2	9	0.44	0.89	0.00
2018	Tier 2	8	0.88	0.88	0.62
2019	Tier 2	11	0.91	1.00	0.36
2020	Tier 2	12	0.67	0.83	0.25
2011	Tier 1	1	1.00	1.00	0.00
2012	Tier 1	2	0.50	1.00	0.50
2013	Tier 1	6	0.33	0.67	0.00
2014	Tier 1	2	0.00	0.00	0.00
2016	Tier 1	1	1.00	0.00	0.00
2017	Tier 1	1	1.00	1.00	1.00
2018	Tier 1	2	1.00	1.00	1.00
2019	Tier 1	3	1.00	1.00	0.67
2020	Tier 1	4	1.00	1.00	0.25
2018	SIDS	1	1.00	1.00	0.00
2019	SIDS	1	1.00	1.00	1.00

readr::write_csv(t, "../output/2w_sids_prop_references.csv")

---

x. Appendix - Total number of regerences by HDI categories (Health Care Sector)

Plot

## j. Appendix - Total number of regerences by HDI categories
p <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::group_by(Sector, Year, hdi_level) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  tidyr::pivot_longer(cols = -c("Sector", "Year", "hdi_level"),
                      names_to = "Type", values_to = "Count") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_references" ~ "Climate",
                                              Type == "intersection_references" ~ "Intersection",
                                              Type == "health_references" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L),
                hdi_level = factor(hdi_level, levels = c("Very high human development",
                                                         "High human development",
                                                         "Medium human development",
                                                         "Low human development"))) %>%
  dplyr::filter(Key == "Intersection" & !is.na(hdi_level)) %>%
  ggplot(aes(x = Year, y = Count, color = hdi_level)) +
  geom_path() +
  theme_minimal() +
  labs(y = "Total number of references",
       color = "") 
p

ggsave("../output/2x_hdi_number_references.pdf", p,  width = 10, height = 7)

Table

total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::group_by(Sector, Year, hdi_level) %>%
  dplyr::summarize(climate_references = sum(climate_count, na.rm = T),
                   health_references = sum(health_count, na.rm = T),
                   intersection_references = sum(intersection_count, na.rm = T)
  ) %>%
  dplyr::filter(!is.na(hdi_level)) %>%
  dplyr::arrange(hdi_level, Year) 

colnames(t) <- c("Sector", "Year", "HDI Level", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Sector	Year	HDI Level	Climate	Health	Intersection
2011	Tier 2	3	0.67	1.00	0.00
2012	Tier 2	6	0.67	0.83	0.17
2013	Tier 2	11	0.36	0.64	0.18
2014	Tier 2	10	0.50	0.80	0.20
2015	Tier 2	12	0.50	0.75	0.08
2016	Tier 2	8	0.50	0.50	0.12
2017	Tier 2	9	0.44	0.89	0.00
2018	Tier 2	8	0.88	0.88	0.62
2019	Tier 2	11	0.91	1.00	0.36
2020	Tier 2	12	0.67	0.83	0.25
2011	Tier 1	1	1.00	1.00	0.00
2012	Tier 1	2	0.50	1.00	0.50
2013	Tier 1	6	0.33	0.67	0.00
2014	Tier 1	2	0.00	0.00	0.00
2016	Tier 1	1	1.00	0.00	0.00
2017	Tier 1	1	1.00	1.00	1.00
2018	Tier 1	2	1.00	1.00	1.00
2019	Tier 1	3	1.00	1.00	0.67
2020	Tier 1	4	1.00	1.00	0.25
2018	SIDS	1	1.00	1.00	0.00
2019	SIDS	1	1.00	1.00	1.00

readr::write_csv(t, "../output/2x_hdi_number_references.csv")

---

y. Appendix - Proportion references by HDI categories (Health Care Sector)

Plot

p <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, hdi_level) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  tidyr::pivot_longer(cols = -c("Year", "hdi_level"),
                      names_to = "Type", values_to = "Prop") %>%
  dplyr::mutate(Key = factor(dplyr::case_when(Type == "climate_prop" ~ "Climate",
                                              Type == "intersection_prop" ~ "Intersection",
                                              Type == "health_prop" ~ "Health"), 
                             levels = c("Intersection", "Climate", "Health")),
                Year = lubridate::ymd(Year, truncated = 2L),
                hdi_level = factor(hdi_level, levels = c("Very high human development",
                                                         "High human development",
                                                         "Medium human development",
                                                         "Low human development"))) %>%
  dplyr::filter(Key == "Intersection" & !is.na(hdi_level)) %>%
  ggplot(aes(x = Year, y = Prop, color = hdi_level)) +
  geom_line() +
  theme_minimal() +
  labs(y = "Proportion of companies, %",
       color = "") +
  scale_y_continuous(labels = scales::percent_format(), limits = c(0, 1))

p

ggsave("../output/2y_hdi_prop_references.pdf", p,  width = 10, height = 7)

Table

t <- total_counts %>% 
  dplyr::filter(Year >= 2011 & Sector == "Health Care Equipment & Ser...") %>%
  dplyr::mutate(climate_n = ifelse(climate_count >= 1, 1, 0),
                health_n = ifelse(health_count >= 1, 1, 0),
                intersection_n = ifelse(intersection_count >= 1, 1, 0)
                ) %>%
  dplyr::group_by(Year, hdi_level) %>%
  dplyr::summarize(total_counts = n(),
                   climate_prop = round(sum(climate_n, na.rm = T)/total_counts,2), 
                   health_prop = round(sum(health_n, na.rm = T)/total_counts,2),
                   intersection_prop = round(sum(intersection_n, na.rm = T)/total_counts,2)
  ) %>%
  dplyr::arrange(desc(hdi_level), Year) %>%
  dplyr::filter(!is.na(hdi_level)) 

colnames(t) <- c("Year", "HDI Level", "Total documents", "Climate", "Health", "Intersection")

t %>%
  knitr::kable() %>%
  kableExtra::kable_styling()

Year	HDI Level	Total documents	Climate	Health	Intersection
2011	Very high human development	9	0.67	1.00	0.00
2012	Very high human development	13	0.54	0.92	0.15
2013	Very high human development	22	0.36	0.86	0.18
2014	Very high human development	21	0.43	0.81	0.14
2015	Very high human development	17	0.65	0.94	0.24
2016	Very high human development	22	0.55	0.82	0.18
2017	Very high human development	26	0.65	0.88	0.12
2018	Very high human development	26	0.85	0.96	0.42
2019	Very high human development	29	0.83	0.97	0.41
2020	Very high human development	37	0.70	0.89	0.41
2011	Medium human development	4	0.25	1.00	0.00
2012	Medium human development	2	0.50	1.00	0.50
2013	Medium human development	3	0.00	1.00	0.00
2014	Medium human development	5	0.20	1.00	0.20
2015	Medium human development	6	0.17	1.00	0.17
2016	Medium human development	4	0.25	1.00	0.25
2017	Medium human development	3	0.00	1.00	0.00
2018	Medium human development	3	0.00	1.00	0.00
2019	Medium human development	4	0.00	1.00	0.00
2020	Medium human development	2	0.50	1.00	0.00
2016	Low human development	1	0.00	1.00	0.00
2017	Low human development	1	0.00	1.00	0.00
2018	Low human development	1	0.00	1.00	0.00
2019	Low human development	2	0.00	1.00	0.00
2020	Low human development	1	0.00	1.00	0.00
2011	High human development	1	1.00	1.00	0.00
2012	High human development	3	0.33	0.67	0.33
2013	High human development	7	0.43	0.43	0.14
2014	High human development	4	0.00	0.50	0.00
2015	High human development	7	0.29	0.57	0.00
2016	High human development	5	0.60	0.40	0.20
2017	High human development	4	0.50	0.50	0.00
2018	High human development	7	0.71	0.57	0.29
2019	High human development	7	0.71	0.86	0.14
2020	High human development	6	0.83	0.83	0.33

readr::write_csv(t, "../output/2y_hdi_prop_references.csv")